totm nov 2023 SpaceX Discussion Thread

[Crimeo]

14 minutes ago, Scotius said:

And yet Delta IV Heavy is three Delta IV boosters tied together. Same goes for Falcon Heavy. Apparently it's easier and cheaper to combine three identical rockets and call it a day, than to build another rocket with higher individual capabilities.

You're talking about the design of things to carry LARGER payloads.

This discussion is about designs for things to carry SMALLER payloads.

Because the falcon heavy already does not recover its first stage when lifting its heaviest payload. It crashes it like normal. It only does the recovery thing when carrying less than maximum payload. So the alternative approach (changing tank size) would be changing it to be smaller, not larger. The options you're talking about of strapping more versions of it to the side thus do not apply: the main reason for that is to have space on the bottom for more engines to lift that extra fuel (versus one bigger tank scales up in volume faster than surface area), but when we are talking about slightly reducing the size, that's not a problem or consideration, really. You can just straight up make it smaller and it should work (pending avionics adjustments, which again are an issue either way).

The manufacturing to do so is almost necessarily cheaper. It's pretty much inconceivable as far as I'm concerned that a 90% length version of the same exact thing is ever going to be harder to make than a completely unrelated new product (like a robot boat).

Thus, it comes down to "is the extra size saved + recovery effort not spent more or less expensive than the first stage"?? And that one is much less intuitively obvious. Thus, was just asking if there's math.

Edited May 3, 2017 by Crimeo

[KSK]

6 hours ago, Crimeo said:

And just like high end sports cars, these sort of assembly lines are by nature a bit "boutique" anyway, with hands-on manual labor used to double check things, etc. Not the same as a toothpick assembly line. See all the ladders and stuff next to the parts, for example.

I think you've hit the nail on the head right there. SpaceX are trying to make their assembly line less boutique to streamline production and (presumably) control costs. So they're going for two 1st stage rocket cores (standard F9 core plus F9H centre core) which they can use in various configurations, plus a common upper stage.

Or at least they were according to this interview with Gwynne Shotwell. That was in 2015, so things may have changed since then but this snippet from the interview still makes interesting reading.

 

Quote

 

Revised Design for Falcon Heavy

Shotwell said the Falcon Heavy will comprise three Falcon 9 core stages, though the central stage will be more robust than the boosters on either side. “Falcon Heavy is two different cores, the inner core and then the two side boosters, and the new single stick Falcon 9 will basically be a Falcon Heavy side booster. So, we're building two types of cores and that's to make sure we don't have a bunch of different configurations of the vehicle around the factory. I think it will streamline operations and really allow us to hit a cadence of one or two a month at every launch site we have.”

 

 

 

[Scotius]

If SpaceX really needed\wanted small launcher for light stuff, they would probably keep Falcon 1 alive. Or even Falcon 9 1.0. I guess they decided to go big 18-wheeler route, and leave small scale "van delivery" stuff to other companies.

[Hannu2]

2 hours ago, Crimeo said:

Thus, it comes down to "is the extra size saved + recovery effort not spent more or less expensive than the first stage"?? And that one is much less intuitively obvious. Thus, was just asking if there's math.

Math is naturally a trade secret of SpaceX and they do not certainly publish it in current business situation. Even they certainly do not know everything yet. Product development is always a risk business. There may always be unforeseeable problems (and also benefits).

But I do not think that you can achieve any significant savings by shortening the tanks. What do you actually try to save? Few tens of meters of welding, hundreds of square meters of metal sheet and some cubic meters of propellants. Cost of those things are very very small part of total costs. The most expensive parts are the rocket engines. They need very advanced machinery and expensive materials. If you can save and reuse them you save more than whole tanks.

Was it ULA or Ariane who research the possibility to return just engines and catch them with helicopter. Reusing was also only reason why main engines were attached on Shuttle instead of external fuel tank.

[RCgothic]

Falcon 9 is now flying heavier payloads reusable than it initially did expendable.

Maybe it might save a little material cost to build a shorter tank for lighter payloads, but you almost certainly wouldn't save more than half the price of a complete booster, as it appears is possible with reusability. Most of the cost is in the engines and if you cut down on mass too much you get over-engined. Redesigning that octaweb is non-trivial. Also, there are additional tooling and inventory management concerns that would eat into your savings.

Finally, Falcon 9 went reuseable because Elon Musk wants to go to mars and things like supersonic retropropulsion and reusability are seen as necessary technologies. I'm convinced he'd go for this direction anyway even if it weren't totally financially optimal.

[wumpus]

10 minutes ago, RCgothic said:

Falcon 9 is now flying heavier payloads reusable than it initially did expendable.

Maybe it might save a little material cost to build a shorter tank for lighter payloads, but you almost certainly wouldn't save more than half the price of a complete booster, as it appears is possible with reusability. Most of the cost is in the engines and if you cut down on mass too much you get over-engined. Redesigning that octaweb is non-trivial. Also, there are additional tooling and inventory management concerns that would eat into your savings.

Finally, Falcon 9 went reuseable because Elon Musk wants to go to mars and things like supersonic retropropulsion and reusability are seen as necessary technologies. I'm convinced he'd go for this direction anyway even if it weren't totally financially optimal.

I'm fairly convinced that the only possible way Spacex could claim to save money on Falcon 9 launches through recovery is by convincing his rocket scientists and engineers to work harder/cheaper for a "cool" goal (there were obviously plenty of savings, and Musk does convince his people to work almost as obsessively as he does).  If the original falcon 9 could be used (even vaguely) in "heavy" configuration, it would be enough to support the much heavier masses currently being sent to orbit.  Presumably adding COTS solid boosters to the side would be enough for the current "block 5" cargoes.

Launching "flight proven" Falcon 9 block 5 will certainly save a ton of money, but will it ever match all the R&D that it took to go from an orbiting Falcon 9 to the final design of block 5?  It  will make plenty of sense getting the same process to go much faster with the raptor (when the costs are vastly higher), but doing so just for Falcon 9 is a harder sell.

[sevenperforce]

13 hours ago, CSE said:

What will be really interesting is whether an F9 expendable or an FH reusable is chosen once both are flying, and if and when there's a payload of the relevant size. I do wonder whether the delay in the FH is due in part to the success of the F9 upgrades to date: the next F9 launch, later this month, was apparently originally intending to use the heavy.

SES-10 was also originally intended to launch on Falcon Heavy, but they managed to put it on F9...and not only F9, but an ASDS-recoverable F9. It's a testament to how much improvement has been made since v1.0.

12 hours ago, Crimeo said:

The structural and aerodynamic load, avionics, center of mass, etc.ALSO change if you're only burning 90% of your fuel, similarly to changing that 10% by stretching or shrinking, and then having the shell fly on a completely different trajectory and do a completely different thing (land on boat, land on pad, splash down, etc.). So those really don't seem very convincing arguments to me. You're going to have to recalibrate all that stuff, and I'm sure it's difficult and expensive and all, but you have to do it with BOTH strategies anyway.

Structural and aerodynamic load change from payload to payload on ascent, so that sort of variability is already factored in. Variance in payload, fuel consumption, and stuff like that is already adjusted in real-time based on exhaustive modeling, so it doesn't really matter whether you're lofting a heavier payload or you're lofting a lighter payload with more fuel reserved for RTLS. In contrast, changing the actual dry mass of the vehicle and altering the overall shape completely wrecks your aerodynamic modeling, along with a bunch of other stuff.

9 hours ago, DVDRW said:

So when and why F9 gets sooty? In video looks like sootyness comes right after separation.

I saw this too; it already looked sooty by the end of the boostback burn. That surprised me; I was expecting most of the soot to be deposited during the re-entry hypersonic retropropulsion phase.

5 hours ago, Scotius said:

If SpaceX really needed\wanted small launcher for light stuff, they would probably keep Falcon 1 alive. Or even Falcon 9 1.0. I guess they decided to go big 18-wheeler route, and leave small scale "van delivery" stuff to other companies.

Or Falcon 5.

It just makes way more sense to reserve extra propellant for a more gentle recovery.

[tater]

The problem with reuse and cost is that they have a staff that needs to be working, and if they are making rockets, then need to launch them. All of them, plus the reused ones. Otherwise they end up with a hanger full of rockets. It comes down to launch cadence as always. Jury is out.

[softweir]

I think the main reason for not building a shorter Falcon, may be economic: if you build a shorter booster for lighter payloads, then that is all it can carry. You lose flexibility. Boosters take months to manufacture and you have to budget to build a booster well in advance. If you have a portfolio of short and long boosters then you find yourself having to second-guess how many of each to build, which multiplies the complexity of your advance planning, budgeting, marketing and sales. Get it wrong and you lose out big time; go to the effort to make sure you get it right adds hugely to the management and budgeting expenses. Those additional expenses would greatly exceed the cost saved (if any!) through building the smaller, "more economical" boosters.

On the other hand, if you build a large number of identical boosters then all you need do is be prepared to put a bit less fuel in any booster that is launching a smaller payload. Yes, using the larger booster wastes fuel and if you fail to recover it then you waste a booster, but the fuel used in a launch is the least expensive item in the cost, and the saving from failing to recover a smaller booster would be offset by the additional costs of deciding when you need to build a new one to replace it.

With their one-size-fits-all Falcons, SpaceX have flexibility and economy of scale. They can launch all but the heaviest of payloads without having to worry about which size of booster to use next.

[Streetwind]

Randomly, because I literally just put two and two together in my head a minute ago:

Remember how we were wondering about whether NROL-76 was vertical or horizontal integration? And how it turned out to be horizontal, despite the Air Force certification for national security launches that SpaceX obtained specifically stipulates vertical integration capability as a hard requirement?

Well, turns out, when you add up the timelines, there is no possible way SpaceX could even have procured this contract after achieving certification. The only explanation is that this wasn't a licensed national security launch at all.

After people picked up on this and poked the NRO enough, they actually came out and confirmed it: this launch was commercially procured. It went to Ball Aerospace for integration, who in turn went and bought a Falcon 9 to do the launch job. And that's how SpaceX got a NRO launch on their manifest before being certified for national security launches: because this one didn't need certification in the first place. It's not a super sensitive thing. People are fairly confident at this point that it's a simple commsat, though one used exclusively by the NRO to talk to its actual spy satellites.

And since it was just a normal commsat, it could be horizontally integrated like any other commsat. Mystery solved!

[TheEpicSquared]

3 minutes ago, softweir said:

snip

Those are very good points.  

[Crimeo]

8 hours ago, Hannu2 said:

Math is naturally a trade secret of SpaceX and they do not certainly publish it in current business situation. Even they certainly do not know everything yet. Product development is always a risk business. There may always be unforeseeable problems (and also benefits).

But I do not think that you can achieve any significant savings by shortening the tanks. What do you actually try to save? Few tens of meters of welding, hundreds of square meters of metal sheet and some cubic meters of propellants. Cost of those things are very very small part of total costs. The most expensive parts are the rocket engines. They need very advanced machinery and expensive materials. If you can save and reuse them you save more than whole tanks.

Was it ULA or Ariane who research the possibility to return just engines and catch them with helicopter. Reusing was also only reason why main engines were attached on Shuttle instead of external fuel tank.

I realize engines are obviously more valuable than a few meters of tank, but are not necessarily obviously more valuable than a few meters of tank PLUS the opportunity benefit of not spending money on robot barges, fancy avionics in the first stage, and landing hardware.

[_Augustus_]

8 hours ago, Hannu2 said:

Was it ULA or Ariane who research the possibility to return just engines and catch them with helicopter?

Both. For Vulcan and Adeline respectively.

[TheEpicSquared]

https://arstechnica.com/information-technology/2017/05/spacexs-falcon-9-rocket-will-launch-thousands-of-broadband-satellites/

2019! Sooner than I expected.  

[Elthy]

Well, 2019 in Elon time. For us mortals that propably means 2021...

[Scotius]

4000+ sats in LEO. With more to come... eventually. Mr. Kessler is sweating bullets at the mere thought.

[Spaceception]

A full satellite network by 2024?

Coincidence? I think not!

They want a network up for the Mars colonists!

I guess that gives another incentive for going to another planet for those who want to go, but still want to look at cat pics, you still receive internet

[softweir]

Be ironic if a Mars launch ran into one of Elon Musk's comsats. 

[CatastrophicFailure]

35 minutes ago, Scotius said:

4000+ sats in LEO. With more to come... eventually. Mr. Kessler is sweating bullets at the mere thought.

Aw, c'mon, now. SpaceX may have a little trouble with dates but their ability to hit what they want to hit is second to none. Therefore one may also assume that their ability to not hit what they don't want to hit is also on par.  

And as someone who lives out on the Styx with limited internet options, this sounds great! Bet the receiver would look real nice next to that PowerWall I'll probably never have, either. 

[Brent Kerman]

5 minutes ago, CatastrophicFailure said:

And as someone who lives out on the Styx

Umm... Styx?

[sevenperforce]

Remembering Falcon 5...

If SpaceX wanted to do a cheaper expendable Falcon 9, the fastest way for them to do so would be to mount fewer engines on the octaweb. They can drop the center engine or any pair of outer engines without affecting balance. Then, simply fill the tanks a little less full than normal, and launch.

Does this make sense? Absolutely not. But it would cost them less than building an entirely new shorter stage and still throwing away all nine Merlins.

[Wjolcz]

Well, I feel like having too much fuel during/after full stage recovery makes much more sense than throwing away the same, except a bit shorter rocket. Even if you build a few-million dollar boat to land the rocket on, it's got to save monies in the long run. Not like it explodes  or sinks after each recovery, or something.

Because that's what's being discussed here right now, isn't it? Having infrastructure and cutting costs later on vs not having infrastructure and cutting costs on expendable rockets?

Edited May 3, 2017 by Veeltch

[sevenperforce]

As an example: I get 23.2 tonnes to a 185x185, 28.5 degree LEO orbit, flying F9 expendable in its current incarnation.

With eight engines (core engine removed), I get 22.7 tonnes.

With seven engines (one pair of engines removed), I get 21.2 tonnes.

With six engines (core and one pair removed), I get 20.5 tonnes.

With five engines (two pairs removed), TWR has dropped low enough that propellant loading needs to be reduced. Reducing to 330 tonnes of fuel on the first stage so it can get off the pad, I get 18.0 tonnes.

If an expendable Falcon 5 could still get more payload to orbit than Falcon 9 flying reusable, it doesn't make sense to build a second smaller "mini-Falcon-9" for expendable payloads. Just reserve the landing propellant and be done with it.

[CatastrophicFailure]

1 hour ago, Brent Kerman said:

Umm... Styx?

Yeah, I live on an island. Pay the ferryman tree fiddy to take me ashore when he's not busy schlepping dead people and epic heroes back & forth. 

[KSK]

46 minutes ago, CatastrophicFailure said:

Yeah, I live on an island. Pay the ferryman tree fiddy to take me ashore when he's not busy schlepping dead people and epic heroes back & forth. 

Don't pay the Ferryman.

Don't even fix a price.

Don"t pay the Ferryman.

Until he gets you to the other side...